Archives
June 2001 Issue
Optical Node Segmentation: Avoiding Future Upgrade Headaches
By Tony Ghaffari, C-COR.net
By using segmentation techniques to upgrade nodes and convert amplifiers to nodes, an operator's initial investment and the cost of future upgrades are minimized.
Planning system upgrades, rebuilds and new builds must be based on forward bandwidth, and on the ability of the reverse path to handle the increasing demand for interactive services, video, Internet, telephony and network management. Proper equipment selection also may mean the difference between thousands of dollars in unnecessary expenses with huge headaches, or a very cost-effective and simple procedure.
An operator should consider many factors when planning an upgrade or a new build, including forward and reverse bandwidth; power allocation; subscriber and service penetration rates; bandwidth per service; and homes per node.
Cable operators have reduced typical node sizes from about 10,000 homes per node when fiber optic technology initially was introduced to about 500 homes per node today. This figure will likely continue to change as demand for services increases. Scalable nodes--with the inherent flexibility for both low or high density, and low or high service demand--are designed to accommodate higher demand for services. As service areas become more dense and as demand increases, an operator simply adds more optical modules to these nodes to accommodate growth.
When we hear the term "optical segmentation," many things come to mind. For the purpose of this article, we will discuss only two forms of optical segmentation:
- Upgrading a node amplifier with additional transmitters and receivers; and
- Installing fiber to an existing amplifier and converting the amplifier to a node.
Many manufacturers offer the options discussed in this article. Some even allow for the easy integration of digital return into their nodes.
Upgrading a node
The primary limiting factor once the fiber backbone has been implemented is the return path. What does an operator do when additional return services are needed, but no more free bandwidth in the return path remains? What does an operator do when the penetration rate of existing return services increases, and return traffic slows to a level subscribers will not tolerate? One solution is upgrading an optical node with additional return transmitters, and segmenting the return path.
It is almost impossible to predict where return services will have high penetration rates. Building a network capable of supporting the maximum anticipated saturation of return services is very costly and inefficient because you will be building small nodes where few or no subscribers live, where little return traffic exists. Installing additional transmitters to segment return services in existing nodes frees valuable bandwidth for both the addition of future return services and faster return traffic.
Forward segmentation also may be required when the need arises for different channel lineups within the same node. For example, let's assume that a retirement community is developed and is served by the same optical node as a military base. Instead of installing a second optical node to serve the retirement community, you may segment the forward path in the existing optical node by adding a second forward receiver. Each receiver may have a separate channel lineup and may feed two outputs of the existing optical node.
A truly segmentable node should allow the forward and return segmentation of each of its ports. This means that a four-port node may be upgraded to accommodate four transmitters and four receivers, each dedicated in its respective path to each port. Using the four-port example, a node with one receiver and one transmitter may be designed and activated to support 2,400 homes passed. As your subscriber penetration increases, you may add three additional return transmitters (and three additional forward receivers) to reduce the number of homes passed to 600. From a financial standpoint, you need to purchase only one node. This node may be upgraded and segmented by purchasing only the needed receivers, transmitters and plug-in accessories. The cost of this upgrade is a fraction of the cost of buying three other complete nodes.
Another option is to add redundancy to your segmented node. In our four-port example, we could dedicate one transmitter or receiver to two ports, and a second transmitter or receiver could be used for redundancy. The redundancy feature also may be an upgrade.
Forward segmentation
Four basic forward configurations exist:
- Single receiver (1 x 4)
- Dual receivers with redundancy or splitband/narrowcasting (2 x 4)
- Dual receivers with segmentation (2 x 2)
- Quad receivers with segmentation (4 x 4)
1 x 4 forward
In the 1 x 4 forward configuration, one fiber feeds optical signals to one forward receiver. The forward signals then are distributed to the radio frequency (RF) module, where they are split to four outputs (see Figure 1).
2 x 4 forward
The 2 x 4 forward configuration (see Figure 2) is a variation of the 1 x 4 configuration and has two versions--one for redundancy and one for splitband/narrowcasting applications. In the 2 x 4 configuration for redundancy, two fibers carry the same signals to two receivers. You control which is the main receiver and which is the redundant backup through an A/B switch. The A/B switch then switches from the main receiver to the backup receiver if the main receiver fails or loses the input signal. The forward signals pass through the A/B switch and are then routed to the RF module, where they are split to four outputs.
In the 2 x 4 configuration for splitband/narrowcasting applications, two fibers carry a different frequency band of signals, each to one of two receivers. The main fiber carries signals in the 50 MHz to 295 MHz bandwidth. The secondary fiber carries signals in the 301 MHz to 862 MHz bandwidth. These signals are combined by a diplex filter and then are routed to the RF module, where they are split to four outputs.
2 x 2 forward
In 2 x 2 forward segmentation (see Figure 3), two fibers carry the same channel lineup or different channel lineups to two receivers. Each set of signals is routed to the RF module and split to two outputs.
Redundancy may be added with additional receivers and A/B switches. In the 2 x 2 configuration for redundancy, a pair of fibers carries the same signals to two receivers, and another pair of fibers carries the same signals to two other receivers. The advantage here is that you may have different channel lineups on each pair of fibers. You control which is the main receiver and which is the redundant backup through an A/B switch. The A/B switch then switches from the main receiver to the backup receiver if the main receiver fails or loses the input signal. The forward signals pass through the A/B switch and then are routed to the RF module, where they are split to two outputs (see Figure 4).
4 x 4 forward
In 4 x 4 forward segmentation (see Figure 5), four fibers carry the same channel lineup or different channel lineups to four receivers. Each set of signals is routed to the RF module and directly to its own output.
Return segmentation
Five basic return configurations exist:
- Single transmitter (4 x 1)
- Dual transmitters with redundancy (4 x 2)
- Dual transmitters with segmentation (2 x 2)
- Dual transmitters with segmentation and redundancy (2 x 4)
- Quad transmitters with segmentation (4 x 4)
4 x 1 reverse
In the 4 x 1 reverse configuration (see Figure 6), the return signals from each port are combined in the RF module and routed to one transmitter via one fiber optic cable to the headend.
4 x 2 reverse
The 4 x 2 reverse configuration (see Figure 7) is a variation of the 4 x 1 configuration, but with redundancy. The return signals from each port are combined in the RF module and then distributed through a reverse splitter that splits the RF signals evenly into two return paths.
Each path then is routed to a separate return transmitter using a separate fiber optic cable running toward the headend.
2 x 2 reverse segmentation
In reverse segmentation, return path operation is similar to that in the 4 x 1 or 4 x 2 reverse configurations, but the possibilities and the complexity begin to increase. Return RF signals from one or more of the four return ports may be segmented, or picked off independently, from the other ports and distributed to one or more return transmitters. In 2 x 2 reverse segmentation (see Figure 8), the return RF signals from two ports are combined and distributed to one return transmitter while the return RF signals from the other two ports are combined and distributed to a second return transmitter. Each set of signals then is sent via a separate fiber optic cable toward the headend.
2 x 4 reverse segmentation
A variation of 2 x 2 reverse segmentation is 2 x 4 reverse segmentation (see Figure 9). Here, two more return transmitters are added for redundancy. Each set of combined signals from two return ports is distributed through a reverse splitter that splits the RF signals evenly into two return paths. Each path then is routed to a separate return transmitter via a separate fiber optic cable toward the headend. Switching for the redundancy occurs in the headend.
4 x 4 reverse segmentation
In 4 x 4 reverse segmentation (see Figure 10), each return port is segmented independently for direct routing to one of four return transmitters. The operator selects which port will be segmented to which return transmitter. This configuration yields the most return bandwidth.
Other more complex reverse segmentation configurations are possible, but these options may have similarities with the reverse path operation described previously. For example, if you have a three-active output node, you may independently segment each return path to one of three-return transmitters.
Or, if you have a four-output node, you may choose to have a three return transmitter configuration where the return signals from two ports are combined and routed to one return transmitter; the return signals from a third port are segmented independently to a second return transmitter; and the return signals from a fourth port are segmented independently to a third return transmitter. The possibilities are numerous and depend on system needs.
Converting an amplifier to a node
In this architecture, an operator divides the basic cable television architecture with a central headend and long amplifier cascades into smaller service nodes.
This division is achieved by strategically upgrading certain amplifiers within the cascades into optical nodes to create a fiber backbone. The existing coaxial cable is then used for shorter RF distribution cascades after the nodes (see Figure 11). This figure is a simplified example of the system.
If service area growth increases beyond what you are able to handle by adding one additional optical node, further expansion is necessary.
For this application, the existing architecture remains essentially the same. The fiber enclosure, or hub, may house an optical splitting network in which spare fibers are stored. Once the decision to upgrade an amplifier to a node has been made, it is just a matter of running fiber from the upgraded node to the fiber enclosure. You may upgrade to more nodes as needed. The only limitation will be the number of spare fibers installed from the headend to the fiber enclosure. You may then activate the dark fibers that were housed in the fiber enclosure and run them to the new nodes. Now, each node passes 125 homes and has shorter distribution cascades.
Aside from the obvious financial advantages (because the new node uses the same RF module and housing), downtime during installation is greatly reduced and different plug-in accessories (pads and equalizers) don't need to be purchased.
Ideally, these are concepts that best serve the interests of the operator when they are part of the design and equipment vendor selection process. If upgradeable nodes and amplifiers are chosen for a new build or rebuild, implementation and cost of segmentation is reduced drastically.
Tony Ghaffari is manager of applications engineering and training for C-COR.net, and is an SCTE International Ambassador. You may reach him at .
Saving Through Scalability
How will a network designed today support tomorrow's bandwidth requirements? One answer lies in the type of amplifiers and nodes deployed. By installing amplifiers that may be upgraded easily to nodes, and nodes that may accept additional transmitters and receivers, an operator will save many future upgrade headaches. The cost of upgrading later is decreased, and the upgrade is easier to justify because you are spending capital where customer demand exists.
With the development of scalable architectures, the need to build a broadband network for "worst case" demand is gone. Today a broadband network may be built for 2,400 homes passed per node. As long as enough spare fibers exist, that node may be segmented down to 300 or even 150 homes passed. Make sure the amplifiers and nodes being installed in your networks are designed to be upgraded and segmented.
|
Back to June 2001 Issue

Access Intelligence's CABLE GROUP
Communications Technology | CableFAX Daily | CableFAX's CableWORLD | CT's Pipeline
CableFAX Magazine | CableFAX databriefs | Broadband Leaders Retreat | CableFAX Leaders Retreat
|